WO1995032556A1 - Montage permettant de determiner le signal de spectre d'un signal numerique a large bande pour en deriver une information d'allocation de bits - Google Patents

Montage permettant de determiner le signal de spectre d'un signal numerique a large bande pour en deriver une information d'allocation de bits Download PDF

Info

Publication number
WO1995032556A1
WO1995032556A1 PCT/IB1995/000348 IB9500348W WO9532556A1 WO 1995032556 A1 WO1995032556 A1 WO 1995032556A1 IB 9500348 W IB9500348 W IB 9500348W WO 9532556 A1 WO9532556 A1 WO 9532556A1
Authority
WO
WIPO (PCT)
Prior art keywords
signal
subband
digital audio
wideband digital
audio signal
Prior art date
Application number
PCT/IB1995/000348
Other languages
English (en)
Inventor
Leon Maria Van De Kerkhof
Pope Ijtsma
Original Assignee
Philips Electronics N.V.
Philips Norden Ab
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Philips Electronics N.V., Philips Norden Ab filed Critical Philips Electronics N.V.
Priority to DE69522883T priority Critical patent/DE69522883T2/de
Priority to JP53016795A priority patent/JP3830106B2/ja
Priority to EP95915999A priority patent/EP0707761B1/fr
Publication of WO1995032556A1 publication Critical patent/WO1995032556A1/fr

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B1/00—Details of transmission systems, not covered by a single one of groups H04B3/00 - H04B13/00; Details of transmission systems not characterised by the medium used for transmission
    • H04B1/66—Details of transmission systems, not covered by a single one of groups H04B3/00 - H04B13/00; Details of transmission systems not characterised by the medium used for transmission for reducing bandwidth of signals; for improving efficiency of transmission
    • H04B1/667—Details of transmission systems, not covered by a single one of groups H04B3/00 - H04B13/00; Details of transmission systems not characterised by the medium used for transmission for reducing bandwidth of signals; for improving efficiency of transmission using a division in frequency subbands

Definitions

  • the invention relates to an arrangement for determining a signal spectrum of a wideband digital audio signal and for deriving bit allocation information in response thereto, in an adaptive bitallocation system, the arrangement comprising
  • - an input terminal for receiving the wideband digital audio signal
  • - signal splitting means for splitting the wideband digital audio signal into M narrow band sub signals, each one of the M sub signals being representative of a component of the wideband digital audio signal which is present in a corresponding one of M adjacent substantially non-overlapping narrow bands in the frequency band of the wideband digital audio signal
  • - calculating means for each time determining from the M sub signals, information which is representative of the signal spectrum of the wideband digital audio signal
  • bit allocation information determining means for deriving bit allocation information in response to the information which is representative of the signal spectrum of the wide band digital audio signal, the bit allocation information being representative of the number of bits with which samples of the sub signals will be represented, and where M is an integer larger than one.
  • bit allocation information being representative of the number of bits with which samples of the sub signals will be represented, and where M is an integer larger than one.
  • the powers thus obtained are processed in a matrix multiplication step so as to obtain masked power values.
  • Those masked power values result in bitneeds b j to b M for the samples in the time equivalent signal blocks of the M subband signals.
  • those bitneed values are used so as to allocate bits to the samples, resulting in the bitallocation information values n l to n M , n m indicating the number of samples with which the 12 samples in the signal block of subband m will be represented, after having carried out a quantization on the samples in the subbands.
  • the invention aims at providing an alternative way of determining the information representative of the signal spectrum of the wideband digital audio signal.
  • the arrangement in accordance with the invention is characterized in that the calculating means comprises transformation means for carrying out on each of the sub signals a time-to-frequency transform based signal processing so as to obtain said information repre ⁇ sentative of the signal spectrum of the wideband digital audio signal.
  • the invention is based on the following recognition.
  • a time-to-frequency transform such as a discrete Fourier transform
  • a time-to-frequency transform such as a discrete Fourier transform
  • Dl discrete Fourier transform
  • D2 discrete Fourier transform
  • a 512- point digital Fourier transform can be carried out on the wideband digital audio signal so as to obtain 257 frequency samples in the frequency range of interest, that is: between 0 Hz and 24 kHz, assuming that the sampling frequency is 48 kHz.
  • Carrying out a time-to-frequency transform based signal processing on the subband signals may result in simpler calculations to be carried out, as the bandwidth of the subbands, and thus the bandwidth of the subband signals is small.
  • the time-to-frequency transform based signal processing such as a Fourier transform based signal processing carried out, is equivalent to carrying out a p-point discrete Fourier transform on the subband signals, where p is a very low number, resulting in a low number of calculations to be carried out in the time-to-frequency transform based signal processing.
  • the circuit realization of the arrangement in accordance with the invention can be much simpler, and thus cheaper.
  • the value q 1 may be chosen larger than 12, the value 12 being the number (q) of values of the signal blocks on which the quantization is carried out, see the prior art documents listed above. More specifically, in the case that q j is chosen larger than 12, the last 12 values in the signal block of q x values may be the signal block of 12 samples on which the quantization will be carried out.
  • the calculating means being adapted to determine each time from M time equivalent signal blocks, one block in each of the M sub signals, each signal block comprising q : samples of a sub signal, said information which is representative of the signal spectrum of the wideband digital audio signal, where q j is an integer larger than one, is further characterized in that the transfor ⁇ mation means being further adapted to carry out on each of the time equivalent signal blocks said time-to-frequency transform based signal processing so as to obtain said information representative of the signal spectrum of the wideband digital audio signal.
  • a time-to-frequency based signal processing is recurrently carried out on a signal block of q- samples in a subband signal, so as to recurrently obtain said information representative of the signal spectrum of the wideband digital audio signal. It may be clear that subsequent signal blocks of q, samples in a subband signal will overlap in the case that q j is larger than q, where q is the number of subsequent signal blocks in a subband signal on which the quantization is carried out.
  • the arrangement in accordance with the invention may be further characterized in that, the transformation means are adapted to carry out a time-to-frequency transform based signal processing which is equivalent to carrying out a time-to-frequency transform on each signal block so as to obtain transform coefficients for each signal block of a subband signal in a subband, deriving therefrom information representative of the signal spectrum in said subband, and combining the information representative of the signal spectra in the M subbands so as to obtain said information representative of the signal spectrum of the wideband digital audio signal.
  • a time-to-frequency transform based signal processing is carried out which is equivalent to one time-to-frequency transform being carried out on the q, samples in a signal block.
  • transfor ⁇ mation means are adapted to carry out a time-to-frequency transform based signal processing which is equivalent to carrying out, on each signal block, n times a time-to-frequency transform on samples of a signal block so as to obtain n sets of transform coefficients for a signal block in a subband, deriving therefrom information representative of the signal spectrum in said subband, and combining the information representative of the signal spectra in the M subbands so as to obtain said information representative of the signal spectrum of the wideband digital audio signal.
  • a time-to-frequency transform based signal processing is carried out which is equivalent to a number of n time-to-frequency transforms being carried out on a signal block.
  • the time-to-frequency transform carried out will preferably be a discrete
  • DCT discrete cosine transform
  • figure 1 shows an embodiment of the arrangement
  • figure 2 shows the subband signals generated by the subband splitter of the arrangement of figure 1
  • figure 3a an embodiment of the bitneed determining means of the arrangement of figure 1
  • figure 3b a further elaborated version of the bitneed determining means.
  • figure 4 shows the power values in a subband
  • figure 5 shows the signal spectra for the various subband signals
  • figure 6 shows the signal spectrum obtained
  • figure 7 shows an embodiment of the calculation unit for carrying out the fourier transform based signal processing.
  • Fig. 1 shows an embodiment of the arrangement.
  • An input terminal I of the arrangement is supplied with a wide-band digital signal.
  • the audio signal may be a stereo audio signal. In that case only one of the two signal portions (the left or right signal portion) of the stereo audio signal will be further discussed. The other signal portion may then be subjected to the same process.
  • Input 1 is supplied with, for example, 16-bit samples of, for example, the left signal portion of the audio signal having a 48 kHz sample frequency.
  • the audio signal is applied to a sub-band splitter 2.
  • the sub-band splitter 2 distributes the audio signal over M sub-bands by means of M filters i.e.
  • a low-pass filter LP M-2 bandpass filters BP and a high-pass filter HP.
  • M is equal to, for example, 32.
  • the sample frequency of the M sub- band signals is reduced in the blocks referenced 9. In such a block the sample frequency is reduced by a factor of M.
  • the signals thus obtained are presented at the outputs 3.1 , 3.2, ... 3.M.
  • the signal is presented in the lowest sub-band SB j .
  • the signal is presented in the lowest but one sub-band SB 2 .
  • the output 3.M the signal is presented in the highest sub-band SB m .
  • the signals at the outputs 3.1 to 3.M have the form of successive samples expressed in 16-bit numbers or more, for example, 24-bit numbers.
  • the sub-bands SB j to SB- ⁇ are all equally wide.
  • a quantizer Q m the q samples of a signal block of the subband signal SB m are quantized to quantized samples having a number of bits n--- smaller than 16.
  • the quantized samples in the sub-bands SB j to SB j ⁇ are then presented at the respective outputs 4.1 to 4.M.
  • the outputs 3.1 to 3.M are furthermore coupled to the respective inputs 5.1 to 5.M of bit need determining means 6.
  • the bit need determining means 6 determines for time-equivalent q- ⁇ -sample signal blocks of the left sub-band signal portion in the sub-bands SB ! to SB M the bit need b m .
  • the bit need b m is a relative number which bears a relationship to the number of bits with which the q samples in a q-sample signal block in a sub-band signal should be quantized.
  • the bit needs b j to b M derived by the bit need determining means 6, are applied to bit allocation means 7.
  • the bit allocation means 7 determines the actual number of bits n, to n M with which the q samples of the corresponding signal blocks in the sub-band signals SB j to SB M are to be quantized on the basis of the bit needs b j to b .
  • Control signals corresponding to the numbers n- ⁇ to n M are applied to the respective quantizers Q j to QJ Y J over the lines 8.1 to 8.M, so that the quantizers are capable of quantizing the samples with the correct number of bits.
  • the quantized samples in the signal blocks of the sub-band signals are thereafter applied to terminals 4.1 to 4.M.
  • the bit allocation information formed from the numbers n l to n M is applied to terminals 12.1 to 12. M.
  • the scale factor information formed from the scale factors SF, to SF M is applied to terminals 11.1 to 11. M.
  • the signal components present at the terminals 4.1 to 4.M, 11.1 to l l.M and 12.1 to 12. M are further processed so as to enable transmission of the said signal components. Reference is made in this respect to document D6.
  • the signal blocks of q ⁇ samples applied to the bitneed determining means 6 may be the same as the signal blocks of q samples on which quantization is carried out. In that case, this results in q being equal to q l . It may however also be possible to take q- larger than q. In that situation, the signal block of q samples to be used for quantization in the quantizing means Q m forms a part of the signal block of q* ⁇ samples to be used in the bitneed determining means 6. More specifically, the signal block of q samples may be the last q samples in the signal block of q j samples. In this situation, subsequent signal blocks of q ⁇ samples in a subband signal will thus overlap in time.
  • Figure 2 schematically shows the subband signals SB- ⁇ to SB M , as they are generated by the subband splitter 2 and supplied to the outputs 3.1 to 3. M respectively.
  • the subband signals are divided into subsequent signal blocks of q ⁇ samples S j to s - each.
  • Processing in the bitneed determining means 6 is each time carried out on time equivalent signal blocks of q j samples. That are those signal blocks included between the vertical broken lines in figure 2. In this situation, it is assumed that q, equals q.
  • Figure 3a shows a more detailed description of the arrangement in accordance with the invention, which is included in the block 6 of figure 1.
  • the block 6 comprises a calculation unit 20 for each time determining from the M subband signals, information which is representative of the signal spectrum of the wideband digital audio signal applied to the terminal 1.
  • the calculation unit 20 has M inputs coupled to the M inputs 5.1 to 5. M of the block 6 so as to receive the M subband signals.
  • the information representative of the signal spectrum of the wideband digital audio signal is available at a number of outputs indicated 21.1, 21.2, 21.3, ... and so on. Those outputs are coupled to corresponding inputs of a further processing unit 32.
  • the information available at the outputs of the unit 20 can be in the form of the M values v-.
  • v M being the signal power or energy of the signal portion or signal block of the subband signal SB m in subband m.
  • the powers v m thus obtained are supplied to the further processing unit 32, which carries out a matrix manipulation, as described in the documents (Dl) and (D2) so as to obtain the bitneeds b ⁇ to b M . It may however also be possible to have more power values than the M power values given above. It will be clear that, in that case, another manipulation on the power values should be carried out in the further processing unit 32 to obtain the M values for the bitneeds b m .
  • the calculation unit 20 is adapted to carry out a discrete Fourier transform based signal processing on each of the subband signals.
  • the said Fourier transform based signal processing is equivalent to carrying out on each of the subband signals, more specifically each of the corresponding time equivalent signal blocks of the subband signals, a discrete Fourier trans ⁇ form (DFT) so as to obtain a set of Fourier transform coefficients for each one of said subband signals, deriving information representative of the signal spectrum in a subband and combining the sets of information so as to obtain the information representative of the signal spectrum of the wideband digital audio signal.
  • the information representative of the signal spectrum in a subband can be in the form of a set of power values for corresponding frequencies in said subband.
  • the said M sets of power values are combined so as to obtain a combined set of power values representing the signal spectrum of the wideband digital audio signal.
  • the Fourier transform based signal processing is equivalent to n times carrying out a discrete Fourier transform, such as a fast Fourier transform, on each of the time equivalent signal blocks of the subband signals.
  • a discrete Fourier transform such as a fast Fourier transform
  • those signal portions in a signal block of a subband signal on which two subsequent Fourier transforms are carried out may in time partly overlap.
  • Each Fourier transform carried out on a signal block results in a set of Fourier coefficients from which a set of power values representing the signal spectrum in the subband can be derived.
  • n sets of such power values are obtained for each subband.
  • Those n sets of power values are combined so as to obtain one combined set of power values for each subband.
  • the M sets of combined set of power values are combined so as to obtain a set of power values representing the signal spectrum of the wideband digital audio signal.
  • the Fourier transform based signal processing is 'equivalent to' carrying out a Fourier transform on a signal portion so as to obtain Fourier transform coefficients and deriving therefrom power values, means that the signal processing need not actually carry out said subsequent steps of carrying out a Fourier transform, obtaining the Fourier transform coefficients and deriving therefrom the power values.
  • the actual signal processing carried out is 'based on' said steps.
  • the signal processing carried out is based on a calculation in which all those steps are subsequently carried out, but where the signal processing only includes the resulting calculation, so that the separate steps on which the calculation is based is not visible anymore in the signal processing carried out.
  • This second embodiment can be realized by the circuit block diagram of figure
  • 3b which shows the calculation unit 20 comprising a transformation unit 22, which generates the M sets of power values (one set for each subband) at outputs 26.1, 26.2, 26.3, ... and so on.
  • Those M sets of power values are supplied to a combining unit 24, to combine the M sets of power values so as to obtain one combined set of power values representing the signal spectrum of the wideband digital audio signal.
  • Y[0] x[0] + x[l] + x[2] + x[3]
  • Y[l] x[0] - j.x[l] - x[2] + j.x[3]
  • Y[2] x[0] - x[l] + x[2] - x[3]
  • Y[3] x[0] * ⁇ * j.x[l] - x[2] - j.x[3]
  • the samples x[0] to x[3] are in the first one of the three fourier transforms carried out on a signal block equal to the samples S j to s 4 respectively.
  • the samples j.[0] to x[3] are in the second one of the three fourier transforms carried out on a signal block equal to the samples s 5 to s 8 respectively.
  • the samples x[0] to x[3] are in the third one of the three fourier transforms carried out on a signal block equal to the samples s 9 to s 12 respectively.
  • the computation explained above is carried out three times for each subband, namely on the first four samples of a signal block, the second four samples in the said signal block and on the third four samples in the signal block.
  • three sets of powers P(0,m), P(l,m) and P(2,m) are obtained.
  • the three values P(0,m) are added so as to obtain the power P(0,m) used for the further processing.
  • the three values P(l,m) are added so as to obtain the power P(l,m) used for the further processing.
  • the three values P(2,m) are added so as to obtain the power P(2,m) used for the further processing.
  • Another combination could be to realize some weighting on the three values for each of the P(0,m), P(l ,m) and P(2,m).
  • the three power values thus obtained for each subband are supplied to the outputs 26. 1 , 26.2, 26.3, .... and so on of the transformation unit 22.
  • the value f s /2 equals 750 Hz. This for the reason that the wideband digital audio signal can have a frequency of 24 kHz, which results in a sampling frequency of 48 kHz, in accordance with the Nyquist sampling theorem. As M is 32, this results in 750 Hz (24000/32) wide subbands.
  • P(0, 1) for this subband thus relates to the actual power of the wideband digital audio signal at a frequency of 0 Hz
  • P(2, l) for this subband thus relates to the actual power of the wideband digital audio signal at a frequency of 750 Hz.
  • the frequency value f s /4 relates to the central frequency of the subband from 750 to 1500 Hz, that is 1125 Hz.
  • the subband from 750 Hz and 1500 Hz not only has been mixed down to the frequency band between 0 Hz and 750 Hz, but has also been mirrored around the central frequency of the subband.
  • the power P(0,2) obtained as given above now has a relation to the power of the wideband digital audio signal at a frequency of 1500 Hz
  • the power P(l ,2) relates to the actual power of the wideband digital audio signal at a frequency of 1125 Hz
  • P(0,3) for this subband thus relates to the actual power of the wideband digital audio signal at a frequency of 1500 Hz
  • P(l,3) for this subband thus relates to the actual power of the wideband digital audio signal at a frequency of 1875 Hz
  • P(2,3) for this subband thus relates to the actual power of the wideband digital audio signal at a frequency of 2250 Hz.
  • this subband runs from 2250 Hz to 3000 Hz.
  • P(0,4) for this subband thus relates to the actual power of the wideband digital audio signal at a frequency of 3000 Hz
  • P(l,4) for this subband thus relates to the actual power of the wideband digital audio signal at a frequency of 2250 Hz
  • P(2,4) for this subband thus relates to the actual power of the wideband digital audio signal at a frequency of 2625 Hz.
  • Figure 5 shows the results for the various subbands, as a function of frequency, present at the outputs of the transformation unit 22.
  • the top line of power spectra are the power spectra for the subbands with odd sequence numbers, running from 1 and up, and the lower line of power spectra are the power spectra for the subbands with the even sequence numbers.
  • Figure 7 shows schematically an embodiment of the calculation unit 20 for carrying out the signal processing described above.
  • the calculation unit 20 comprises the transformation unit 22 and the signal combining unit 24 of figure 3b.
  • the transformation unit 22 comprises memories 50.1 to 50. q, of which only the memories 50.m-l , 50. m and 50.m + l are shown.
  • the memories have four storage locations for storing four subsequent sample values indicated x[0] to x[3].
  • the memories 50.1 to 50. M have inputs coupled to the inputs 5.1 to 5.
  • M of the bit determining unit 6 for receiving the subband signals. Only the inputs 5.m-l , 5. m and m-r- 1 are shown in figure 7.
  • the memory 50.m-l stores each time four sample values of the subband signal SB m.1 . As has been said previously, those four sample values can be the sample values s l to s 4 , or the sample values s 5 to s 8 or the sample values s 9 to s 12 °f a signal block of the subband signal SB. ⁇ .
  • the memory 50. m stores each time four sample values of the subband signal SB m .
  • the four sample values stored in the memory 50. m are time equivalent to the samples stored in the memory 50.m-l.
  • the memory 50.m+ l stores each time four sample values of the subband signal SB m + 1 .
  • the four sample values stored in the memory 50.m+ l are time equivalent to the four samples stored in the memories 50.m-l and 50. m.
  • the calculation block 52.1 is adapted to calculate the power P(0,m) from the four sample values x[0] to x[3] in accordance with the equation given above.
  • the calculation block 52.2 is adapted to calculate the power P(l,m) from the four sample values x[0] to x[3] in accordance with the equation given above.
  • the calculation block 52.3 is adapted to calculate the power P(2,m) from the four sample values x[0] to x[3] in accordance with the equation given above.
  • Identical calculation blocks are present for each of the other memories 50.1 to
  • Figure 7 shows the calculation block 53 for calculating the power P(2,m+ 1) from the sample values x[0] to x[3] stored in the memory 50.m-!- l, and the calculation block 54 for calculating the power P(0,m-1) from the sample values x[0] to x[3] stored in the memory 50.m-l.
  • the outputs of all calculation blocks are coupled to a first input of an adder unit, such as the adder units 56.1 , 56.2 and 56.3.
  • An output of the adder units is coupled to an input of a memory, such as the memories 58.1 , 58.2 and 58.3.
  • Outputs of the memories are coupled to a second input of the adder units.
  • the outputs of the memo ⁇ es 59 and 58.1 are further coupled to conesponding inputs of an adder unit 62 and the outputs of the memories 58.3 and 60 are coupled to conesponding inputs of an adder unit 63.
  • the signal processing on time equivalent signal blocks is as follows.
  • the memories 58.1 , 58.2, 58.3, 59 and 60 are cleared so that their contents is zero.
  • the first four sample values S j to s 4 of all the time equivalent signal blocks of the M subband signals are stored in the memories 50.1 to 50.M.
  • the three power values P(0,m), P(l.m) and P(2,m) are calculated in the calculation blocks 52.1, 52.2 and 52.3. Those three power values supplied to the adders 56. 1 , 56.2 and 56.3.
  • P(0,m), P(l ,m) and P(2,m) for a subband signal SB m are calculated and supplied to the respective adders 56.1 , 56.2 and 56.3.
  • the power values now calculated are added to the power values stored in the memories 58.1, 58.2 and 58.3 and the added values are again stored in the said memories.
  • the third set of four sample values s 9 to s 12 of all the time equivalent signal blocks are stored in the memories 50.1 to 50. M.
  • the three power values P(0,m), P(l,m) and P(2,m) for a subband signal SB m are calculated and supplied to the respective adders 56.1, 56.2 and 56.3.
  • the power values now calculated are added to the power values stored in the memories 58.1 , 58.2 and 58.3 and the added values are again stored in the said memories.
  • the power value now stored in the memory 58.2 is the power value P 2m , see above. Further, the power values P(0,m-1) and P(0,m) stored in the memories 59 and 58.1 respectively are added together in the adder unit 62 so as to obtain the power value P 2m . l , see also above. In the same way, the power values P(2,m) and P(2,m-t- l) stored in the memories 58.3 and 60 respectively are added together in the adder unit 63 so as to obtain the power value P 2m+ 1 , see also above.
  • the signal combination unit 24 thus comprises a number of adder units, such as the adder units 62 and 63 shown in figure 7, for adding each time the power value tor the lowest frequency in a subband with the power value for the highest frequency in the next lower subband.
  • the multiplier 65 has its input coupled to the output of the memory 64, in which the power value P.0.1 ) is stored.
  • the multiplier 65 multiplies P(0, 1) by 2 so as to obtain P 0 .
  • a windowing such as a Hamming windowing
  • an aliasing compensation well known in the art is preferably carried out in rhe case when p j is larger than 4.

Abstract

Montage permettant de déterminer le signal de spectre d'un signal numérique audio à large bande dans un système adaptatif d'allocation de bits et comprenant un moyen (2) de partage dudit signal audio en M sous-signaux à bande étroite (SBm) dont chacun est représentatif d'une composante du signal audio à large bande présent dans l'une des M bandes étroites correspondantes (contiguës mais ne se recouvrant pratiquement pas) de la bande de fréquence du signal audio numérique à large bande. Il comporte en outre un moyen de calcul (20) de la durée de chaque sous-signal M, ces informations étant représentatives du spectre du signal audio à large bande, et M étant un entier supérieur à un. Le moyen de calcul (20) comporte un moyen de transformation (22) permettant d'effectuer sur chacun des sous-signaux un traitement du signal à l'aide d'une transformée temps/fréquence afin d'obtenir ladite information représentative du spectre du signal audio numérique à large bande. La transformée temps/fréquence peut être de Fourier.
PCT/IB1995/000348 1994-05-19 1995-05-10 Montage permettant de determiner le signal de spectre d'un signal numerique a large bande pour en deriver une information d'allocation de bits WO1995032556A1 (fr)

Priority Applications (3)

Application Number Priority Date Filing Date Title
DE69522883T DE69522883T2 (de) 1994-05-19 1995-05-10 Vorrichtung zur bestimmung des signalspektrums eines digitalen breitbandsignals und zur ableitung einer bitzuweisungsinformation
JP53016795A JP3830106B2 (ja) 1994-05-19 1995-05-10 広帯域ディジタル信号の信号スペクトルを規定し、かつこの信号スペクトルに応じてビット配置情報を得る機構
EP95915999A EP0707761B1 (fr) 1994-05-19 1995-05-10 Montage permettant de determiner le signal de spectre d'un signal numerique a large bande pour en deriver une information d'allocation de bits

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
EP94201427 1994-05-19
EP94201427.5 1994-05-19

Publications (1)

Publication Number Publication Date
WO1995032556A1 true WO1995032556A1 (fr) 1995-11-30

Family

ID=8216890

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IB1995/000348 WO1995032556A1 (fr) 1994-05-19 1995-05-10 Montage permettant de determiner le signal de spectre d'un signal numerique a large bande pour en deriver une information d'allocation de bits

Country Status (5)

Country Link
US (1) US5784533A (fr)
EP (1) EP0707761B1 (fr)
JP (1) JP3830106B2 (fr)
DE (1) DE69522883T2 (fr)
WO (1) WO1995032556A1 (fr)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100261254B1 (ko) * 1997-04-02 2000-07-01 윤종용 비트율 조절이 가능한 오디오 데이터 부호화/복호화방법 및 장치
US5903872A (en) * 1997-10-17 1999-05-11 Dolby Laboratories Licensing Corporation Frame-based audio coding with additional filterbank to attenuate spectral splatter at frame boundaries

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0457391A1 (fr) * 1990-05-14 1991-11-21 Koninklijke Philips Electronics N.V. Méthode de codage et dispositif de codage comportant un codeur de sous-bande et émetteur muni du dispositif de codage
US5150387A (en) * 1989-12-21 1992-09-22 Kabushiki Kaisha Toshiba Variable rate encoding and communicating apparatus
US5260980A (en) * 1990-08-24 1993-11-09 Sony Corporation Digital signal encoder

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
NL8700985A (nl) * 1987-04-27 1988-11-16 Philips Nv Systeem voor sub-band codering van een digitaal audiosignaal.
US5222189A (en) * 1989-01-27 1993-06-22 Dolby Laboratories Licensing Corporation Low time-delay transform coder, decoder, and encoder/decoder for high-quality audio
NL9000338A (nl) * 1989-06-02 1991-01-02 Koninkl Philips Electronics Nv Digitaal transmissiesysteem, zender en ontvanger te gebruiken in het transmissiesysteem en registratiedrager verkregen met de zender in de vorm van een optekeninrichting.
EP0400222A1 (fr) * 1989-06-02 1990-12-05 ETAT FRANCAIS représenté par le Ministère des Postes, des Télécommunications et de l'Espace Système de transmission numérique utilisant le codage par sous-bandes d'un signal numérique
NL9001127A (nl) * 1990-05-14 1991-12-02 Philips Nv Kodeerinrichting bevattende een subbandkoder, en een zender voorzien van de kodeerinrichting.
US5367608A (en) * 1990-05-14 1994-11-22 U.S. Philips Corporation Transmitter, encoding system and method employing use of a bit allocation unit for subband coding a digital signal
US5365553A (en) * 1990-11-30 1994-11-15 U.S. Philips Corporation Transmitter, encoding system and method employing use of a bit need determiner for subband coding a digital signal
ATE138238T1 (de) * 1991-01-08 1996-06-15 Dolby Lab Licensing Corp Kodierer/dekodierer für mehrdimensionale schallfelder

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5150387A (en) * 1989-12-21 1992-09-22 Kabushiki Kaisha Toshiba Variable rate encoding and communicating apparatus
EP0457391A1 (fr) * 1990-05-14 1991-11-21 Koninklijke Philips Electronics N.V. Méthode de codage et dispositif de codage comportant un codeur de sous-bande et émetteur muni du dispositif de codage
US5260980A (en) * 1990-08-24 1993-11-09 Sony Corporation Digital signal encoder

Also Published As

Publication number Publication date
US5784533A (en) 1998-07-21
JP3830106B2 (ja) 2006-10-04
DE69522883T2 (de) 2002-04-11
DE69522883D1 (de) 2001-10-31
EP0707761A1 (fr) 1996-04-24
EP0707761B1 (fr) 2001-09-26
JPH09501037A (ja) 1997-01-28

Similar Documents

Publication Publication Date Title
JP4242450B2 (ja) フェーザ測定用単一側波帯フィルタバンクの効率的実施装置
JP3134455B2 (ja) 高能率符号化装置及び方法
JP3093179B2 (ja) 高品質オーディオ用短時間遅延変換エンコーダ及びデコーダ
KR100253136B1 (ko) 저계산 복잡도의 디지탈 필터뱅크
Shlien Guide to MPEG-1 audio standard
JP2904472B2 (ja) ディジタル・オーディオ信号を効率的に圧縮するための方法、データ処理システムおよび装置
EP0799531B1 (fr) Procede et appareil pour appliquer une prediction des formes d'onde a des sous-bandes d'un systeme de codage perceptif
PL174314B1 (pl) Sposób i urządzenie do dekodowania sygnałów cyfrowych
EP0411655B1 (fr) Codeur et décodeur utilisant la quantification vectorielle
JPH04177300A (ja) 音声帯域分割符号化装置
US5758020A (en) Methods and apparatus for encoding and decoding signals, methods for transmitting signals, and an information recording medium
KR0185998B1 (ko) 인코딩 시스템 및 이를 구비한 트랜스미터
US5754127A (en) Information encoding method and apparatus, and information decoding method and apparatus
EP0697665B1 (fr) Méthode et appareil pour coder, transmettre et décoder des informations
US5850418A (en) Encoding system and encoding method for encoding a digital signal having at least a first and a second digital component
JPH0846518A (ja) 情報符号化方法及び復号化方法、情報符号化装置及び復号化装置、並びに情報記録媒体
JP3065343B2 (ja) 信号伝送方法
EP0707761B1 (fr) Montage permettant de determiner le signal de spectre d'un signal numerique a large bande pour en deriver une information d'allocation de bits
US6064698A (en) Method and apparatus for coding
JPH09102742A (ja) 符号化方法および装置、復号化方法および装置、並びに記録媒体
JP3010637B2 (ja) 量子化装置及び量子化方法
JP4114244B2 (ja) エンコード方法、デコード方法、エンコード装置、デコード装置、ディジタル信号記録方法、ディジタル信号記録装置、ディジタル信号送信方法及びディジタル信号送信装置
JPH07221650A (ja) 情報符号化方法及び装置並びに情報復号化方法及び装置
JP3513879B2 (ja) 情報符号化方法及び情報復号化方法
JP3057258B2 (ja) ディジタルデータの高能率符号化装置及び方法

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): JP

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE

WWE Wipo information: entry into national phase

Ref document number: 1995915999

Country of ref document: EP

121 Ep: the epo has been informed by wipo that ep was designated in this application
WWP Wipo information: published in national office

Ref document number: 1995915999

Country of ref document: EP

WWG Wipo information: grant in national office

Ref document number: 1995915999

Country of ref document: EP